Welfare analysis on optimal enterprise tax rate in China

Economic Modelling 33 (2013) 149–158

Contents lists available at SciVerse ScienceDirect

Economic Modelling journal homepage: www.elsevier.com/locate/ecmod

Welfare analysis on optimal enterprise tax rate in China☆ Jingjing Ji a, 1, Zhiqiang Ye b, 2, Shunming Zhang c,⁎ a b c

Haitong Securities Company Limited, Shanghai 200001, PR China Department of Finance, School of Business, East China University of Science and Technology, Shanghai 200237, PR China China Financial Policy Research Center, Renmin University of China, Beijing 100872, PR China

a r t i c l e

i n f o

Article history: Accepted 21 March 2013 JEL classification: H21 H71 Keywords: General equilibrium Welfare analysis Enterprise tax Taxation reform

a b s t r a c t This paper builds a model of general equilibrium for production economies to analyze Chinese enterprise tax reform which regulated the unified enterprise tax rate to be at 25%. The reform was backed by the new Law on Corporate Income Tax executed from January 1, 2008. Using national statistics of 2007, we obtain that the optimal unified enterprise tax rate for manufacturing industries is 21.82% if tax revenue is given. In addition, we find the globally optimal enterprise tax rates are 33.11%, 18.17%, and 18.06% for state-owned enterprises (SOEs), foreign invested enterprises (FIEs) and other private enterprises (OPEs), respectively. Our results suggest the achieved aim to pack those inefficient FIEs off and gain a competitive edge for China. Comparing the optimal unified enterprise tax rate equilibrium with benchmark equilibrium, unified enterprise tax rate at 25% equilibrium and globally optimal enterprise tax rate equilibrium, we conclude that the optimal unified enterprise tax rate (21.82%) is an efficient policy for Chinese government. At last, it shows the reliability of the conclusion when sensitivity analysis on enterprise tax rate for FIEs and premium coefficient is performed for optimal unified enterprise tax rate equilibrium. © 2013 Elsevier B.V. All rights reserved.

1. Introduction The new law on enterprise tax has been put into practice in China since January 1, 2008. This law unifies the tax rates for domestic (invested) enterprises and foreign invested enterprises (FIEs) which were unequal. Few economists do research theoretically on this problem whether, in whole, this reform of unified enterprise tax rate promotes or inhibits economic growth. China was implementing a series of favorable terms for foreign invested enterprises in order to bring in foreign investment after the policy of reform and opening-up. One of the important terms was to cut down enterprise tax rate into a low level. Owing to those favorable terms, the amount of foreign investments rose steadily year by year and China's FDI (foreign direct investment) reaches 74.8 billion USD in 2007. Not only foreign invested enterprises did bring funds, which improved the situation of shortage of money, but introduced advanced technology and administrative experiences, and supplied large numbers of job offers. All of these highly contributed to China's rapid economic development. However, the super national-wide treatment for ☆ This work is supported by the National Natural Science Foundation of China (NSFC Grant Numbers: 70825003 and 71273271) and the National Social Science Foundation of China (SSFC Grant Number: 07AJL002). ⁎ Corresponding author. Tel.: +86 10 8250 0626. E-mail addresses: [email protected], [email protected] (J. Ji), [email protected] (Z. Ye), [email protected], [email protected] (S. Zhang). 1 Tel.: +86 15102145113. 2 Tel.: +86 21 64253152. 0264-9993/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.econmod.2013.03.019

foreign invested enterprises drove domestic enterprises at a disadvantage, and at the meantime, most of those foreign invested enterprises in China were labor-intensive, or at the middle- and low-end of the industry chain. By right of China's low cost of labor, land and raw materials, foreign invested enterprises made huge profits while negatively affected China's resources and environment. Therefore, the central government started to consider cleaning out the mass environmental pollution and intensive energy consumption in the wake of those low value-added enterprises, and updating the industrial structure for the sake of development of China's economy. Under such circumstances, the Law on Enterprise Tax was adopted at the 5th Session of the Standing Committee of the 10th National People's Congress of the People's Republic of China. The new law unified the enterprise tax rate at 25% and was comprehensively enforced as of January 1, 2008. Wang (1995) investigates the dual tax system which implies different tax rates, as well as different operations on tax deduction and tax base, and concludes that domestic enterprises' tax burden is more severe than foreign enterprises. Chen (2003) finds out that the difference of domestic and foreign enterprises tax burden stems from the different requirements of tax credit and tax rate, and shows empirically that the real tax burden of domestic enterprises is twice as much as that of foreign enterprises. Whalley and Wang (2007) explore the effect of state-owned enterprises on social welfare after the new Law on Enterprise Tax came into force. They build models with a general equilibrium of state-owned enterprises controlled by workers and by managers. Using statistics of 2004, they conclude in the general equilibrium model of state-owned enterprises controlled by workers that there is a 0.26% welfare loss under the new Law of Enterprise Tax, and even

150

J. Ji et al. / Economic Modelling 33 (2013) 149–158

more loss in the optimal tax structure; whereas in the general equilibrium model of state-owned enterprises controlled by managers, the first loss is 0.19%. van der Hoek et al. (2008) survey the effect of the new Law of Enterprise Tax on foreign invested enterprises. They show that implementing of unified tax rate is a necessary reform so that the new law would do no harm on the increment of foreign investments. Moreover, since the increased tax rate of foreign invested enterprises by the new law made them the most progressive ones, the reform of enterprise tax impacted positively on China's economic growth. The existing literature has not answered an important question: what the optimal unified tax rate for manufacturing industries is. Although Whalley and Wang (2007) have discussed the optimal tax rate, to define it they create a case where state-owned enterprises apply one rate and there is another one for foreign invested enterprises and other private enterprises. That means in their paper there are still more than one tax rate. In our paper, however, the optimal unified enterprise tax rate (OUETR) is defined as a unified one for all manufacturing sectors under optimization of maximizing the social welfare. Besides, this paper uses the data of 2007, with the feature of cancelled agricultural tax; while Whalley and Wang (2007) use the data of 2004 which includes agricultural tax. Furthermore, we loose the restriction that different sectors have to share a unique enterprise tax rate to investigate the globally optimal enterprise tax rates (GOETRs). Thus globally optimal enterprise tax rate area is a set of rates rather than a sole one. To put it another way, globally optimal enterprise tax rates are those for corresponding sectors with which the whole economy reaches its optimum. By application of the CGE model, this paper focuses on solving the optimal tax rate through the following steps. In the general equilibrium model, we have four production sectors. They are agricultural sector (Agriculture), state-owned enterprises (SOEs), foreign invested enterprises (FIEs) and other private enterprises (OPEs). The added value of the secondary and tertiary industries is distributed into the last three sectors according to the corresponding share of manufacturing added value. The wage rate and the return of capital are determined by competitive market conditions for those non state-owned enterprises. As to monopolistic state-owned enterprises, the wage rate is assumed to be administratively fixed, and the return to capital is decided by the marginal product with respect to the input since capital flows freely among sectors. Based on the data of 2007, we calibrate the parameters of the production function and the utility function in the model, and obtain the optimal enterprise unified tax rate (OUETR) of 21.82% and globally optimal enterprise tax rates (GOETRs) with the results for SOEs, FIEs and OPEs of 33.11%, 18.17% and 18.06%, respectively with the given tax revenue. To reach the optimum, the labor input moves from agricultural sector to manufacturing sectors, and the capital input moves from agricultural and SOE sectors to FIE and OPE sectors, with the output in agricultural and FIE sectors decreasing and in SOE and OPE sectors increasing. Compared with UETR25-equilibrium (i.e., equilibrium with the unified enterprise tax rate of 25%),3 there are increments on social welfare, national income and expenditure in the equilibrium with the optimal unified tax rate of 21.82%, and we also make a comparison between OUETR-equilibrium (i.e., equilibrium with optimal unified tax rate) 4 and GOETR-equilibrium (i.e., equilibrium with globally optimal enterprise tax rates) 5 to find that these two equilibria are very close. So the results show that the optimal unified tax rate of 21.82% is an efficient

3 The unified enterprise tax rate of 25% was executed for the three production sectors by the new Law on Corporate Income Tax from January 1, 2008. 4 The equilibrium with optimal unified enterprise tax rate implies that the equal enterprise tax rates for the three production sectors ensure utility maximization in equilibrium. 5 The equilibrium with globally optimal enterprise tax rates implies that the enterprise tax rates for the three production sectors attain the maximum utility value in equilibrium.

policy. At last, this paper performs the sensitivity analysis on the optimal unified enterprise tax rate with real enterprise tax rates for foreign invested enterprises and premium coefficient for managers in SOEs varying in arranged intervals. It shows the reliability of the conclusion given tax revenue. The structure of this paper is as follows. We construct a general equilibrium model with production and consumption included in Section 2. Section 3 calibrates the parameters of production function and utility function based on the data set of China's economy in 2007. In Section 4 we investigate the optimal unified enterprise tax rate, and compare and analyze the equilibria of benchmark, of the unified enterprise tax rate of 25% and of the optimal unified enterprise tax rate. Sections 5 and 6 execute the sensitivity analysis as to enterprise tax rate for FIEs and premium coefficient, respectively. Section 7 concludes. 2. Basic model The general equilibrium model will be set up on production economies. We consider a small open price taking economy with four sectors: agriculture, state-owned enterprises (SOEs), foreign investment enterprises (FIEs), and other private enterprises (OPEs). The world prices for the four goods are Pj 0 for j = 0, 1, 2, and 3. Domestic prices are then given by world prices plus (or minus) the effect of ad valorem border measures for import tariffs as well as export subsidies, i.e.
 P j ¼ 1 þ τ j P 0j ; j ¼ 0; 1; 2; and 3

ð2:1Þ

where τj for j = 0, 1, 2, and 3 are agricultural and manufacturing import tariffs or export subsidies (τ0 > 0 indicates import tariffs for agricultural sector, and τj b 0 indicates export subsidies for manufacturing sectors j = 1, 2, and 3). For the four sectors we assume the production function to be of the Cobb–Douglas form α

1−α j

Y j ¼ ϕ j Lj j K j

;

ð2:2Þ

j ¼ 0; 1; 2; and 3

where Yj is the output, Lj and Kj are the labor and capital (factor) inputs, respectively, ϕj is the units term (scalar parameter), and αj is the production exponent. For non-SOE sectors j = 0, 2, and 3, the markets are fully competitive. The wage rate and the return of capital are determined by the marginal products with respect to labor and capital, respectively. W j ¼ Pj

Rj ¼ P j

∂Y j ∂Lj

α −1

¼ P j ϕj α j Lj j

1−α j

Kj

¼ αj

Pj Y j ; Lj

j ¼ 0; 2; and 3;

ð2:3Þ


 α −α
P Y j j ¼ P j ϕj 1−α j Lj j K j j ¼ 1−α j ; j ¼ 0; 2; and 3: Kj ∂K j ∂Y j

ð2:4Þ In this case the zero-profit condition holds. P j Y j ¼ W j L j þ Rj K j

j ¼ 0; 2; and 3:

ð2:5Þ

For the SOE sector, we assume that members in this sector face the wage W1 which is fixed by policy. Besides wages paid to laborers, the SOE enterprises must pay the interest for the capital they receive from state banks. The profit they are entitled to is zero. This yields a budget constraint that the enterprise membership must meet, P 1 Y 1 ¼ W 1 L1 þ R1 K 1 :

ð2:6Þ

Given the economy we investigate on is a small open price taking one, the government can control the level of laborer's effort by

J. Ji et al. / Economic Modelling 33 (2013) 149–158

adjusting SOE wage rate, which brings changes in budget constraint. From this budget constraint, we are also led to infer

R1 ¼

P 1 Y 1 −W 1 L1 : K1

ð2:7Þ

151

amounts of consumption in each sector. The representative consumer receives income from working in four sectors, and this income I is actually gross domestic production value. 3 X

I¼

P j Y j þ B þ S þ D:

ð2:11Þ

j¼0

Harberger (1959, 1962) examines the enterprise tax rate for two production sectors, where incorporated production enterprises have to pay the enterprise tax, while non-incorporated production enterprises must not. In our paper, the agricultural sector is non-incorporated, with the enterprise tax rate t0 = 0, while the SOE, OPE, and FIE sectors are incorporated, with the enterprise tax rate tj > 0 for j = 1, 2, and 3. The above item Rj is the rate of return of gross-of-tax capital. We denote R′j to be the rate of return of net-of-tax capital. Then
 Rj ¼ 1 þ t j R′j ; j ¼ 0; 1; 2; and 3:

B¼

′

R0 ¼

γ

′

′

¼ R2 ¼ R3 :

δ

δ

δ

ð2:10Þ

where δj for j = 0, 1, 2, and 3 are the share parameters with δ0 + δ1 + δ2 + δ3 = 1, and Xj for j = 0, 1, 2, and 3 are the corresponding

0

ð2:12Þ

Net exports exist in manufacturing sectors, then subsidy revenue S is paid for exports, S¼

3   3 X   0 − X 0 τj Pj Z j : τj P j Z j ¼ j¼0

ð2:13Þ

j¼1

As we know, trade imbalance D is defined as D¼

3 X

0 0

ð2:14Þ

Pj Zj

j¼0

and tax revenue V is defined by V¼

3 X

′

t j Rj K j ¼

j¼0

3 X

′

t j Rj K j :

ð2:15Þ

j¼1

Under the budget constraint the representative consumer allocates his income in four kinds of goods, and maximizes his utility. δ

δ

δ

δ

max U ¼ X 00 X 11 X 22 X 33 s:t: P 0 X 0 þ P 1 X 1 þ P 2 X 2 þ P 3 X 3 ¼ I:

ð2:16Þ

Utility maximization subject to a budget constraint in this case implies Xj ¼

U ¼ X 00 X 11 X 22 X 33

0 þ

τj Pj Z j ¼ τ0 P0 Z 0 :

j¼0

ð2:9Þ

Although state-owned enterprises (SOEs) undergo the marketoriented reform, they continue to enjoy some preferential terms and privileges, such as market access, procurability of capital and land-use. The governments (and state-owned asset supervision and administration commissions) prevent the managers of state-owned enterprises from enjoying these preferential terms and privileges and utilizing these capital inputs with low efficiency when the enterprise tax for SOE decreases. So R1′ is set out as the necessary rate of return of capital for state-owned enterprises when other sectors are considered in this model. On the demand side of the model, domestic consumption decision of four sectors reflects utility maximizing behavior by a single representative individual who stands for the aggregation demand in the economy and has a preference of Cobb–Douglas form δ

3 X

ð2:8Þ

In the mid 1990s, state-owned enterprises (SOEs) started a difficult reform in China — many (unskilled) workers got the ax, a lot of firms were contracted, transferred, and auctioned, and so on, and then the share system of modern enterprises was built gradually. From 2000 state-owned enterprises (SOEs) have been out of the woods, which shows that the reform achieved marked results. With economic efficiency improving significantly, the characteristic of managerial control for share system of modern enterprises formed the initial formulation. Therefore Chinese state-owned enterprises are consistent with managerial control SOE model in Whalley and Zhang (2006) and Whalley and Wang (2007). Our paper determines the demand of labor inputs for state-owned enterprises (SOEs) on the condition of marginal product for input. However, the central government announces mandatory requirements for the rate of return of capital input for state-owned enterprises (SOEs). For example, capital inventory for SOEs is assured to be increasing and the managers of SOEs are not allowed to over-invest. Therefore, in equilibrium, the rate of return of net-of-tax capital for SOEs is greater than other sectors. The national capital endowment flows without limitation, which results to the equalized interest rate across the three production sectors excluding SOEs. We have Eq. (2.9) thereby, with γ as the premium coefficient of the return of capital for SOEs with respect to other sectors. R′1

Let Zj denote the amount of net trade, Zj > 0 and Zj b 0 represent net import and net export for sector j. Based on the previous assumptions, the government imposes tariff merely on net import. As to the situation of China's trade, there exists net import only in agricultural sector (j = 0). Therefore, tariff revenue B comes from agricultural imports, i.e.

δj I ; Pj

j ¼ 0; 1; 2; and 3:

ð2:17Þ

A general equilibrium for this model is characterized by the wage rate and the return of capital. Given exogenously a policy parameter (W1) and factor endowments (L, K), the equilibrium is determined by values of (Yj, Lj, Kj, Wj, Rj, Xj), j = 0, 1, 2, and 3 with the following conditions: X j ¼ Y j þ Z j ; j ¼ 0; 1; 2; and 3;

ð2:18Þ

L0 þ L1 þ L2 þ L3 ¼ L;

ð2:19Þ

K 0 þ K 1 þ K 2 þ K 3 ¼ K:

ð2:20Þ

The representative consumer makes his consumption decision under the budget constraint. Zj is the net trade of each sector induced by the difference between production and consumption of each good. Trade balance follows directly from the budget constraint. If reduction in tariff rate lowers domestic prices and raises output of manufacturing sectors from the budget constraint, it is opposite to the scenario in a traditional competitive model in which a lowered

152

J. Ji et al. / Economic Modelling 33 (2013) 149–158

tariff rate reduces output in the protector sector. Since labor in the SOE firms does not receive its marginal product, the equilibrium will not satisfy conditions of Pareto optimality. To assess the impact of tax policy changes on social welfare, this paper applies the conventional indicator, Hicksian Equivalent Variations of Welfare Change (EV). It measures by money and facilitates the evaluation of different policies. EV is defined as: EV = E(U 1, P 0) − E(U 0, P 0). E is the payment function, the least expenditure to acquire utility U under the price level of P. The superscripts 0 and 1 represent the equilibrium before and after policy changes, respectively. EV measures the amount of expenditure that a consumer is required to increase to be as satisfied as after policy changes but under the price level before policy changes. Shoven and Whalley 1 0 0 (1992) show that EV can be simplified as: EV ¼ U U−U I if the utility 0 0 function is linear homogeneous, where I denotes the income before policy changes. Thus, the amplitude of EV relative to the benchmark income is 1

HEV ¼

0

EV U −U  100% ¼  100%: I0 U0

ð2:21Þ

3. Data set, calibration and counterfactual analysis We are employing the formulation set out above to conduct numerical analyses which explore tax policy and calculate the optimal tax rate for manufacturing sectors. We calibrate the model to a 2007 benchmark equilibrium data set capturing the welfare analysis for enterprise tax rate in manufacturing sectors. We draw on data from the Chinese National Bureau of Statistics (NBS, 2008) (the China Statistical Yearbook and China Labour Statistical Yearbook) for our calibrations. The agricultural sector is taken to be the primary industry (farming, forestry, animal husbandry and fishery) in the yearbook. We group the manufacturing sector in China across sectors of registration as reported in the China Statistical Yearbook and allocate each to the three manufacturing sub-sectors appearing in the model of SOEs, FIEs and OPEs. Sectors of registration according to the statistical yearbook are as: Manufacturing 1 = State-Owned Industry, Manufacturing 2 = Collective-Owned Industry, Manufacturing 3 = Co-operative Enterprises, Manufacturing 4 = Joint Ownership Enterprises, Manufacturing 5 = Limited Liability Co-operations, Manufacturing 6 = Share Holding Enterprises, Manufacturing 7 = Private Enterprises, Manufacturing 8 = Other Enterprises, Manufacturing 9 = Enterprises with Funds from Hong Kong, Macao, and Taiwan, and Manufacturing 10 = Foreign Funded Enterprises. We treat the first 5 as the SOE sub-sector, Manufacturing 9 and 10 as the FIE sub-sector, and Manufacturing 6, 7, and 8 as the OPE sub-sector. From the NBS data (Table 1) the 2007 gross domestic products (GDP) of the four sectors are P0Y0 = 2809.5000 billion RMB, P1Y1 = 8649.0387 billion RMB, P2Y2 = 6424.1712 billion RMB, and P3Y3 = 7079.2800 billion RMB; the values of labor earnings are W0L0 = 2543.6953 billion RMB, W1L1 = 3950.7777 billion RMB,

W2L2 = 3776.3851 billion RMB, and W3L3 = 3849.7303 billion RMB; the values of net trade are P00Z0 = 1380.7039 billion RMB, P10Z1 = − 1286.5509 billion RMB, P20Z2 = − 1030.9331 billion RMB, and P30Z3 =− 1054.1450 billion RMB. We refer to literature sources for both import tariffs and export subsidy rates for China's agricultural and manufacturing trades. The export subsidy for manufacturing reflects tax preferences for exports shipped to foreign owned enterprises. The average tariff rate on imports (agricultural good) is τ0 = 15.23%, and the average export subsidy (subsidy rate on the manufacturing good) is τj = − 5.926955% for j = 1, 2, and 3 (from CHINANEWS (2008 and 2009)). The agricultural tax rate is t0 = 0%, and the enterprise tax rate of manufacturing sectors is t1 = 30%, t2 = 15%, and t3 = 22% (from CHINANEWS (2008 and 2009) http://finance.people.com.cn/GB/67610.5215781. html). These four rates on domestic prices are relative to world prices. Following Harberger (1962) and Shoven and Whalley (1972), we define physical units for agricultural and manufacturing products such that in the initial benchmark equilibrium, Pj0 = 1 for j = 0, 1, 2, and 3. The wage rate in agricultural sector is set to be a unit, W0 =1.000000, and the wage rates of manufacturing sectors are obtained by the ratios of their weighted average payment on labor input to agriculture, W1 = 2.290940, W2 = 2.451058, and W3 = 1.776789. The FDI in 2007 is assumed to be the value of capital input in FIEs K2 = 74.8 × 7.604 = 568.7792 billion RMB. In benchmark dataset we set the premium coefficient of the return of capital for SOEs with respect to other sectors to be γ = 1. These data source thus yields a benchmark data set that we are able to use in calibrating our model. From Eq. (2.1) we get the domestic price P0 = 1.152300 and Pj = 0.940730 for j = 1, 2, and 3. Table 1 gives us the production Y0 = 2438.1671, Y1 = 9184.3936, Y2 = 6828.9181, and Y3 = 7525.3012. From Eq. (2.3) or (2.4) we have the production exponents α0 = 0.905391, α1 = 0.457264, α2 = 0.587840, and α3 = 0.543803, which is share of labor earning to GDP. From Table 1 we get the labor input L0 = 2543.6953, L1 = 1724.5225, L2 = 1540.7166, and L3 = 2166.6782; then the total endowment of labor input is L = 7975.6126. From Eqs. (2.5) and (2.6) we get the value of capital input R0K0 = 265.8047 billion RMB, R1K1 = 4689.2610 billion RMB, R2K2 = 2647.7861 billion RMB, and R3K3 = 3229.5497 billion RMB; then we obtain the capital return of FIEs R2 = 4.655209 and R2′ = 4.048008 from Eq. (2.8). From Eq. (2.9) we have R0′ = R1′ = R2′ = R3′ = 4.048008; and from Eq. (2.8) we have R0 = 4.048008, R1 = 5.262410, and R3 = 4.938570. Then we obtain the capital inputs K0 = 65.6631, K1 = 891.0862, and K3 = 653.9443; hence the total endowment of capital input is K = 2179.4728. From Eq. (2.2) we obtain the unit terms (scalar parameters) ϕ0 = 1.354726, ϕ1 = 7.620979, ϕ2 = 6.683491, and ϕ3 = 5.998848. From Table 1 we get Z0 = 1380.7039, Z1 = − 1286.5508, Z2 = − 1030.9331, and Z3 = − 1054.1450. From Eq. (2.18) we have the consumption for the four goods X0 = 3818.8710, X1 = 7897.8428, X2 = 5797.9851, and X3 = 6471.1563. On the other hand, from Eqs. (2.12)–(2.15) we obtain the tariff revenue B = 210.2812, the subsidy revenue S = 199.8349, the trade imbalance D = −1990.9249,

Table 1 Benchmark dataset in China 2007 used in calibrating alternative formulation.

Tariff–subsidy rate Tax rate World price GDP Labor earning Wage rate Capital input Value of net trade

Agriculture

SOEs

τ0 t0 P00 P0Y0 W0L0 W0

τ1 t1 P10 P1Y1 W1L1 W1

= = = = = =

15.23% 0% 1 2809.5000 2543.6953 1.000000

P00Z0 = 1380.7039

FIEs = = = = = =

−5.926955% 30% 1 8640.0387 3950.7777 2.290940

P10Z1 = −1286.5508

τ2 t2 P20 P2Y2 W2L2 W2 K2 P20Z2

OPEs = = = = = = = =

−5.926955% 15% 1 6424.1712 3776.3851 2.451058 568.7792 −1030.9331

τ3 t3 P30 P3Y3 W3L3 W3

= = = = = =

−5.926955% 22% 1 7079.2800 3849.7303 1.776789

P30Z3 = −1054.1450

J. Ji et al. / Economic Modelling 33 (2013) 149–158

and the tax revenue V = 2009.8784; thus the total income is I = 23,372.1812. From Eq. (2.17) we obtain the exponential parameters δ0 = 0.188279, δ1 = 0.317888, δ2 = 0.233369, and δ3 =0.260464. From Eq. (2.10) we obtain the utility value U = 6084.5742. We are to perform numerical simulation analyses for various forms of enterprise tax in China using the model formulation set out above. Using the model parameterizations generated by calibration we then parametrically vary enterprise tax rate and assess the response of the whole economy. This allows us to be aware of the welfare implications of alternative enterprise tax policy changes in China capturing taxation behavioral response using the formulation. We can also compare the results we generate to a comparable competitive case by calibrating a simple competitive model with benchmark equilibrium and general equilibrium with unified enterprise tax rate of 25% using the same data sets. In each experiment we first calibrate the relevant model to the 2007 benchmark data set out in Table 1, and then vary unified enterprise tax rate and compute the relevant new equilibria. Finally, we also obtain globally optimal enterprise tax rate which verifies the reliability of the choice for the unified enterprise tax rate. 4. The optimal enterprise tax rate in China With the aim of maximizing social welfare, this paper obtains the optimal unified enterprise tax rate. Substituting the data in Table 1 and the calibrated parameters in Table 2 into the model and running GAMS software of the equilibrium program, we have benchmark equilibrium in the first column of Table 3. In the benchmark equilibrium, the agricultural tax rate is t0 = 0%, the enterprise tax rates of manufacturing sectors are t1 = 30%, t2 = 15%, and t3 =22%, the wage rates are W0 = 1.000000, W1 = 2.290940, W2 =2.451058, and W3 = 1.776789, the capital returns are R0 =4.048008, R1 = 5.262410, R2 = 4.655209, and R3 = 4.938570 (the net return is R0′ = R1′ = R2′ = R3′ = 4.048008), and the utility value (social welfare) is 6084.5742. We next consider three equilibria with unified enterprise tax rate of 25% (UETR25), with optimal unified enterprise tax rate (OUETR), and with globally optimal enterprise tax rates (GOETR) based on the benchmark equilibrium as follows. As the enterprise tax rates in manufacturing sectors have been unified at 25% since January 1, 2008, we examine general equilibrium with this unified rate. In the equilibrium model we set the wage rate in SOEs fixed, W1 = 2.29094006. When digging the equilibrium, we set the tax revenue to be fixed at V = 2009.87837759. Table 3 reports the equilibrium. We assume that there exists a proportional change in net trade. The equilibrium net trade is assumed to be Zj = θZj 0 for θ ≥ 0, j = 0, 1, 2, and 3. We report the general equilibrium with the united enterprise tax rate of 25% (UETR25-equilibrium) in column 2. Next we consider the optimal united enterprise tax rate. We maintain the assumption that there exists a proportional change in net trade.

153

We fix the tax revenue, V = 2009.87837759, and list the result of the general equilibrium with optimal unified enterprise tax rate (OUETR-equilibrium) in column 3. The optimal unified enterprise tax rate is t = 21.818725%. Finally, we consider the globally optimal enterprise tax rates. To do this, there are variables in need of being fixed: FDI as the capital input in FIE sector, K2 = 568.7792, the tax revenue, V = 2009.87837759, and net trade, Z0 = 1380.7039, Z1 = − 1286.5509, Z2 = − 1030.9331, and Z3 = − 1054.1450. Then we work out general equilibrium with the globally optimal enterprise tax rates (GOETR-equilibrium). The optimal enterprise tax rates are t1 = 33.111791%, t2 = 18.169132%, and t3 = 18.063638%. The results of GOETR-equilibrium are presented in column 4 of Table 3. 4.1. Comparison between benchmark equilibrium and UETR25-equilibrium We now explore if the economic policy of unified enterprise tax rate of 25% is efficient. From benchmark equilibrium (the enterprise tax rates of manufacturing sectors are t1 = 30%, t2 = 15%, and t3 = 22%) to UETR25-equilibrium, the total income and the consumptions decrease by 0.950203%, the social welfare decreases from 6084.5742 to 6026.7584, and the HEV (Hicksian Equivalent Variation) is − 0.950203% (see column 1 in Table 4). The execution of the unified enterprise tax rate of 25% triggers a heavy of outflow of input from agricultural sector with a weakened income and a reduced social welfare. Remember the criterion to judge the tax rate policy is the representative agent's utility. Based on that, the unified enterprise tax rate of 25% leaves much to be desired for Chinese government as well as the whole economy. 4.2. Comparison between benchmark equilibrium and OUETR-equilibrium From benchmark equilibrium to OUETR-equilibrium, the total income and the consumption increase by 0.064447%, the social welfare increases from 6084.5742 to 6088.4955, and the HEV (Hicksian Equivalent Variation) is 0.064447%. For agricultural sector, the labor and capital inputs move to manufacturing sectors by 33.241299% and 38.389293%, respectively, and the output decreases by 34.454800%. For SOE sector, the labor and capital inputs increase by 15.479265% and 3.129013%, respectively, and the output increases by 8.603322%. Unlike the benchmark case, FIE sector requires more labor in OUETRequilibrium even at a more unpleasant cost and less capital inputs in which Chinese government is not thirsty for any more. So part of the FIE firms move to other Asian countries, such as India, Indonesia, Vietnam, and so on, where they can enjoy preferable terms and conditions. Thus the optimal unified enterprise tax rate of 21.818725% is an efficient policy for Chinese government and the whole economy.

Table 2 Calibrated model parameters and benchmark equilibrium (data set).

Domestic price Units term Production exponent Production Labor input Capital input Capital value Return of capital Net trade Consumption Exponential Parameter

Agriculture

SOEs

FIEs

OPEs

P0 = 1.152300 ϕ0 = 1.354726 α0 = 0.905391 Y0 = 2438.1671 L0 = 2543.6953 K0 = 65.6631 R0K0 = 265.805 R0 = 4.048008 R′0 = R′1 = R′2 = R′3 = 4.048008 Z0 = 1380.704 X0 = 3818.8710 δ0 = 0.188279

P1 = 0.940730 ϕ1 = 7.620979 α1 = 0.457264 Y1 = 9184.3936 L1 = 1724.5225 K1 = 891.0862 R1K1 = 4689.261 R1 = 5.262410

P2 = 0.940730 ϕ2 = 6.683491 α2 = 0.587840 Y2 = 6828.9181 L2 = 1540.7166 K2 = 568.7792 R2K2 = 2647.786 R2 = 4.655209

P3 = 0.940730 ϕ3 = 5.998848 α3 = 0.543803 Y3 = 7525.3012 L3 = 2166.6782 K3 = 653.9443 R3K3 = 3229.550 R3 = 4.938570

Z1 = −1286.551 X1 = 7897.8428 δ1 = 0.317888

Z2 = −1030.933 X2 = 5797.9851 δ2 = 0.233369

Z3 = −1054.145 X3 = 6471.1563 δ3 = 0.260464

154

J. Ji et al. / Economic Modelling 33 (2013) 149–158

Table 3 Comparisons among benchmark equilibrium, UETR25-equilibrium, OUETR-equilibrium, and GOETR-equilibrium.

Enterprise tax rate

Domestic price Production

Labor input

Capital input

Wage rate

Capital return

Multiplicity Net trade

Tariff revenue Subsidy revenue Trade imbalance Tax revenue Total income Consumption

Utility value Equivalent variation

Benchmark

UETR25

OUETR

GOETR

Equilibrium

Equilibrium

Equilibrium

Equilibrium

t0 t1 t2 t3

t0 t1 t2 t3

t0 = 0% t1 = 30% t2 = 15% t3 = 22% P0 = 1.15230000 and P1 = P2 = P3 Y0 = 2438.1671 Y1 = 9184.3936 Y2 = 6828.9181 Y3 = 7525.3012 L0 = 2543.6953 L1 = 1724.5225 L2 = 1540.7166 L3 = 2166.6782 K0 = 65.6631 K1 = 891.0862 K2 = 568.7792 K3 = 653.9443 W0 = 1.000000 W1 = 2.290940 W2 = 2.451058 W3 = 1.776789 R0 = 4.048008 R1 = 5.262410 R2 = 4.655209 R3 = 4.938570 θ = 1.000000 Z0 = 1380.7039 Z1 = −1286.5508 Z2 = −1030.9331 Z3 = −1054.1450 B = 210.2812 S = 199.8349 D = −1990.9249 V = 2009.8784 I = 23,372.1812 X0 = 3818.8710 X1 = 7897.8428 X2 = 5797.9851 X3 = 6471.1563 U = 6084.5742

t0 = 0% t1 = 25% t2 = 25% t3 = 25% = 0.94073045 Y0 = 1273.4367 Y1 = 10,160.8405 Y2 = 7616.4027 Y3 = 8325.3603 L0 = 1318.0691 L1 = 2634.3949 L2 = 1727.5227 L3 = 2295.6259 K0 = 36.9967 K1 = 751.1638 K2 = 629.5889 K3 = 761.7234 W0 = 1.007954 W1 = 2.290940 W2 = 2.438095 W3 = 1.855275 R0 = 3.752440 R1 = 4.690550 R2 = 4.690550 R3 = 4.690550 θ = 1.817295 Z0 = 2509.1473 Z1 = −2338.0433 Z2 = −1873.5102 Z3 = −1915.6931 B = 382.1431 S = 363.1591 D = −3618.0993 I = 23,150.0980 X0 = 3782.5840 X1 = 7822.7972 X2 = 5742.8924 X3 = 6409.6671 U = 6026.7584 HEV = −0.950203%

4.3. Comparison between UETR25-equilibrium and OUETR-equilibrium From the two above results, it is obvious that OUETR-equilibrium is much better than UETR25-equilibrium. We report this at column 2 in Table 4. From UETR25-equilibrium to OUETR-equilibrium, both the total income and consumption increase by 0.024384%, the social welfare increases from 6026.7584 to 6088.4955, and the HEV (Hicksian Equivalent Variation) is 0.024384%. For agricultural sector, the labor and capital inputs increase by 27.314289% and 9.348942%, respectively, and the output increases by 25.495166%. For SOE sector, the labor input decreases by 24.405187% and the capital input increases by 22.339282%, and the output decreases by 1.833351%. For FIE sector, the labor input increases by 5.968686% and the capital input decreases by 12.406032%, and the output decreases by 2.030574%. For OPE sector, the labor input increases by 7.832173% and the capital input decreases by 12.229689%, and the output decreases by 1.833352%. So we can also conclude that the unified enterprise tax rate of 21.818725% is efficient. 4.4. Comparison between OUETR-equilibrium and GOETR-equilibrium Column 3 in Table 4 reports that GOETR-equilibrium is better off than OUETR-equilibrium. From equilibrium with optimal unified enterprise tax rate to equilibrium with globally optimal enterprise tax rate

= = = =

0% 21.818725% 21.818725% 21.818725%

= = = =

0% 33.111790% 18.169132% 18.063638%

Y0 = 1598.1015 Y1 = 9974.5566 Y2 = 7461.7460 Y3 = 8172.7271 L0 = 1678.0903 L1 = 1991.4659 L2 = 1830.6331 L3 = 2475.4233 K0 = 40.4555 K1 = 918.9684 K2 = 551.4819 K3 = 668.5670 W0 = 0.993552 W1 = 2.290940 W2 = 2.254050 W3 = 1.688978 R0 = 4.306517 R1 = 5.246144 R2 = 5.246144 R3 = 5.246144 θ = 1.610215 Z0 = 2223.2307 Z1 = −2071.6239 Z2 = −1660.0243 Z3 = −1697.4004 B = 338.5980 S = 321.7773 D = −3205.8179

Y0 = 2441.4063 Y1 = 9191.0924 Y2 = 6833.8359 Y3 = 7530.7900 L0 = 2540.0405 L1 = 1829.8687 L2 = 1542.6045 L3 = 2063.0988 K0 = 67.5136 K1 = 848.8040 K2 = 568.7792 K3 = 694.3759 W0 = 1.002769 W1 = 2.290940 W2 = 2.449821 W3 = 1.867355 R0 = 3.942283 R1 = 5.247643 R2 = 4.658562 R3 = 4.654403 θ = 1.000000 Z0 = 1380.7039 Z1 = −1286.5508 Z2 = −1030.9331 Z3 = −1054.1450 B = 210.2812 S = 199.8349 D = −1990.9249

I = 23,387.2438 X0 = 3821.3322 X1 = 7902.9327 X2 = 5801.7217 X3 = 6475.3267 U = 6088.4955 HEV = 0.064447%

I = 23,392.0052 X0 = 3822.1102 X1 = 7904.5417 X2 = 5802.9029 X3 = 6476.6450 U = 6089.7351 HEV = 0.084819%

(t1 = 33.111791%, t2 = 18.169132%, and t3 = 18.063638%), the social welfare increases from 6088.4955 to 6089.7351, and the HEV (Hicksian Equivalent Variation) is 0.020360%. If we focus on the enterprise tax rate, the weighted average enterprise tax rate based on GDP is 3 X tj Pj Y j j¼0 3 X

¼ 21:408962%, which is very close to OUETR case. This seems Pj Y j

j¼0

an indirect evidence to support OUETR of 21.818725%.

5. Sensitivity analysis on enterprise tax rate for FIEs We analyze the OUETR-equilibrium under the assumption that tax revenue keeps unchanged. The enterprise tax rate for FIE sector would intensively affect the equilibrium results. Besides, it is hard to acquire the exact real tax rate for FIEs on the condition that the favorable rates offered to FIEs are various across the whole country. Consolidating the relevant research work we fix the range of the real enterprise tax rate for FIEs into the interval [0%, 25%]. Adopting the similar analysis steps of Section 4, we list the results in Table 5 concerning the sensitivity effect of the tax rate for FIEs upon equilibrium outcomes. As the tax rate for FIEs rises, the utility value of

J. Ji et al. / Economic Modelling 33 (2013) 149–158

155

Table 4 Equilibria changes from benchmark equilibrium to UETR25-equilibrium, to OUETR-equilibrium, and to GOETR-equilibrium. Benchmark (%)

Equilibrium change from benchmark case to unified tax rate of 25%

Equilibrium change from unified tax rate of 25% to optimal unified tax rate

Equilibrium change from optimal unified tax rate to globally optimal tax rate

Production

Y_ 0 ¼ −47:770737% Y_ 1 ¼ 10:631588% Y_ 2 ¼ 11:531616% Y_ 3 ¼ 10:631589% L_ 0 ¼ −48:182901% L_ 1 ¼ 52:760831% L_ 2 ¼ 12:124624% L_ 3 ¼ 5:951401% K_ 0 ¼ −43:656787% K_ 1 ¼ −15:702454% K_ 2 ¼ 10:691266% K_ 3 ¼ 16:481388% _ 0 ¼ 0:007954% W _ 1 ¼ 0:000000% W _ 2 ¼ 0:528874% W _ 3 ¼ 4:417294% W R_ 0 ¼ −7:301567% R_ 1 ¼ −10:866884% R_ 2 ¼ 0:759171% R_ 3 ¼ −5:022102% Z_ ¼ 81:729573% B_ ¼ 81:729573% S_ ¼ 81:729573% D_ ¼ 81:729573% I_ ¼ −0:950203% X_ 0 ¼ −0:950203% X_ 1 ¼ −0:950203% X_ 2 ¼ −0:950203% X_ 3 ¼ −0:950203% U_ ¼ −0:950203%

Y_ 0 ¼ 25:495166% Y_ 1 ¼ −1:833351% Y_ 2 ¼ −2:030574% Y_ 3 ¼ −1:833352% L_ 0 ¼ 27:314289% L_ 1 ¼ −24:405187% L_ 2 ¼ 5:968686% L_ 3 ¼ 7:832173% K_ 0 ¼ 9:348942% K_ 1 ¼ 22:339282% K_ 2 ¼ −12:406032% K_ 3 ¼ −12:229689% _ 0 ¼ −1:428835% W _ 1 ¼ 0:000000% W _ 2 ¼ −7:548721% W _ 3 ¼ −8:963469% W R_ 0 ¼ 14:765779% R_ 1 ¼ 11:844965% R_ 2 ¼ 11:844965% R_ 3 ¼ 11:844965% Z_ ¼ −11:394971% B_ ¼ −11:394971% S_ ¼ −11:394971% D_ ¼ −11:394971% I_ ¼ 0:024384% X_ 0 ¼ 0:024384% X_ 1 ¼ 0:024384% X_ 2 ¼ 0:024384% X_ 3 ¼ 0:024384% U_ ¼ 0:024384%

Y_ 0 ¼ 52:769164% Y_ 1 ¼ −7:854627% Y_ 2 ¼ −8:415056% Y_ 3 ¼ −7:854625% L_ 0 ¼ 51:364947% L_ 1 ¼ −8:114485% L_ 2 ¼ −15:733825% L_ 3 ¼ −16:656727% K_ 0 ¼ 66:883613% K_ 1 ¼ −7:635127% K_ 2 ¼ 3:136513% K_ 3 ¼ 3:860331% _ 0 ¼ 0:927682% W _ 1 ¼ 0:000000% W _ 2 ¼ 8:685300% W _ 3 ¼ 10:561239% W R_ 0 ¼ −8:457740% R_ 1 ¼ −0:028611% R_ 2 ¼ −11:200264% R_ 3 ¼ −11:279542% Z_ ¼ −37:896508% B_ ¼ −37:896508% S_ ¼ −37:896508% D_ ¼ −37:896508% I_ ¼ 0:020360% X_ 0 ¼ 0:020360% X_ 1 ¼ 0:020360% X_ 2 ¼ 0:020360% X_ 3 ¼ 0:020360% U_ ¼ 0:020360%

Labor input

Capital input

Wage rate

Capital return

Net trade Tariff revenue Subsidy revenue Trade imbalance Total income Consumption

Utility value

benchmark equilibrium maintains at the level of 6084.5742, that of UETR25-equilibrium increases from 6008.9618 to 6036.2438, and that of OUETR-equilibrium decreases from 6096.5958 to 6082.6612, while OUETR increases from 20.650935% to 22.501949%. Fig. 1 describes how the utility values vary across the three equilibria. The utility of benchmark equilibrium is a horizon line at the

level of 6084.5742, that of UETR25-equilibrium is a line with a positive slope of

6036:2438−6008:9618 27:2820 ¼ 25% 25%

Table 5 Sensitivity analysis on enterprise tax rate for FIEs in benchmark equilibrium, UETR25-equilibrium, and OUETR-equilibrium. Tax rate for FIEs (%)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Utility value: benchmark equilibrium

Utility value: UETR25 equilibrium

OUETR equilibrium Utility value

OUETR (%)

6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742

6008.9618 6010.3219 6011.6224 6012.9058 6014.1749 6015.4219 6016.6472 6017.8514 6019.0344 6020.1969 6021.3392 6022.4611 6023.5634 6024.6480 6025.7126 6026.7585 6027.7859 6028.7952 6029.7866 6030.7604 6031.7167 6032.6563 6033.5778 6034.4829 6035.3719 6036.2438

6096.5958 6096.1091 6095.6117 6095.1046 6094.5884 6094.0639 6093.5318 6092.9926 6092.4471 6091.8958 6091.3393 6090.7781 6090.2126 6089.6434 6089.0709 6088.4955 6087.9177 6087.3377 6086.7560 6086.1729 6085.5887 6085.0037 6084.4182 6083.8324 6083.2467 6082.6612

20.650935 20.734554 20.817328 20.899260 20.980353 21.060611 21.140039 21.218644 21.296432 21.373410 21.449585 21.524967 21.599563 21.673382 21.746433 21.818725 21.890267 21.961069 22.031139 22.100488 22.169124 22.237058 22.304298 22.370853 22.436734 22.501949

156

J. Ji et al. / Economic Modelling 33 (2013) 149–158

U 6104

OUETR Equilibrium 6088

Benchmark Equilibrium 6072 6056 6040

UETR25 Equilibrium 6024 6008 O

5%

10%

15%

20%

25% t2

Fig. 1. Sensitivity analysis on enterprise tax rate for FIEs in benchmark equilibrium, UETR25-equilibrium, and OUETR-equilibrium.

from 6008.9618 to 6036.2438, and that of OUETR-equilibrium is a line with a negative slope of 6082:6612−6096:5958 13:9346 ¼− 25% 25% from 6096.5958 to 6082.6612. Thus, Table 5 and Fig. 1 address the following points: (1) As the tax rate for FIEs increases, the utility function of benchmark equilibrium is a horizon line fixed at the level of 6084.5742, that of UETR25-equilibrium is a line with a positive slope (increasing) from 6008.9618 to 6036.2438, and that of OUETR-equilibrium is a line with a negative slope (decreasing) from 6096.5958 to 6082.6612;

(2) No matter how the real tax rate for FIEs varies, utility value of benchmark equilibrium is surely superior to that of UETR25equilibrium; (3) No matter how the real tax rate for FIEs varies, utility value of UETR25-equilibrium is surely inferior to that of OUETRequilibrium; (4) The decreasing utility line of OUETR-equilibrium intersects the horizon utility line of benchmark equilibrium at the real tax rate for FIEs of 21.7336%. Table 6 reports sensitivity analysis on enterprise tax rate for FIEs in GOETR-equilibrium. Routinely, we fix the range of the real enterprise tax rate for FIEs into the interval [0%, 25%]. Utility function is decreasing gradually from 6091.3764 to 6088.7692. As the tax rate for FIEs grows, however, the globally optimal enterprise tax rate

Table 6 Sensitivity analysis on enterprise tax rate for FIEs in GOETR-equilibrium. Tax rate for FIEs (%)

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25

Equilibrium with globally optimal enterprise tax rate Utility value

Enterprise tax rate (%)

6091.3764 6091.2593 6091.1433 6091.0285 6090.9147 6090.8021 6090.6906 6090.5801 6090.4708 6090.3625 6090.2553 6090.1492 6090.0441 6089.9401 6089.8371 6089.7351 6089.6341 6089.5341 6089.4352 6089.3372 6089.2401 6089.1441 6089.0490 6088.9548 6088.8615 6088.7692

t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1 t1

= = = = = = = = = = = = = = = = = = = = = = = = = =

38.365311%, 37.966309%, 37.575150%, 37.191630%, 36.815519%, 36.446607%, 36.084683%, 35.729555%, 35.381033%, 35.038933%, 34.703080%, 34.373305%, 34.049443%, 33.731339%, 33.418836%, 33.111791%, 32.810061%, 32.513509%, 32.222005%, 31.935419%, 31.653628%, 31.376512%, 31.103957%, 30.835850%, 30.572084%, 30.312554%,

t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2 t2

= = = = = = = = = = = = = = = = = = = = = = = = = =

6.852190%, t3 = 23.089262% 7.606471%, t3 = 22.706664 8.360781%, t3 = 22.331746% 9.115120%, t3 = 21.964279% 9.869486%, t3 = 21.604044% 10.623882%, t3 = 21.250826% 11.378298%, t3 = 20.904429% 12.132743%, t3 = 20.564652% 12.887212%, t3 = 20.231308% 13.641705%, t3 = 19.904216% 14.396221%, t3 = 19.583203% 15.150761%, t3 = 19.268089% 15.905322%, t3 = 18.958743% 16.659906%, t3 = 18.654980% 17.414508%, t3 = 18.356661% 18.169132%, t3 = 18.063638% 18.923776%, t3 = 17.775773% 19.678438%, t3 = 17.492930% 20.433112%, t1 = 17.214979% 21.187822%, t3 = 16.941795% 21.942541%, t3 = 16.673255% 22.697276%, t3 = 16.409243% 23.452030%, t3 = 16.149645% 24.206799%, t3 = 15.894351% 24.961585%, t3 = 15.643255% 25.716387%, t3 = 15.396254%

J. Ji et al. / Economic Modelling 33 (2013) 149–158

varies dramatically as follows. The enterprise tax rate for SOE sector decreases from 38.365311% to 30.312554%, for FIE sector is increasing from 6.852190% to 25.716387%, and for OPE sector is increasing from 23.089262% to 15.396254%. 6. Sensitivity analysis on premium coefficient To have managers of SOEs in control, the central government sets a distinct rate of capital return for SOEs from market capital return rate. The premium coefficient is the ratio of capital return rate in SOEs to capital return rate of market. We are rarely able to measure premium coefficient exactly which maybe affect the optimal unified enterprise tax rate. Consolidating the relevant research work we fix the range of premium coefficient into the interval [0.50, 1.50]. Adopting the similar analysis steps of Section 4, we list the results in Table 7 concerning the sensitivity effect of premium coefficient upon equilibrium outcomes. The utility value of benchmark equilibrium maintains at the level of 6084.5742, that of UETR25 equilibrium decreases from 6139.788241 to 5991.405474. The interesting is that the utility value of OUETR-equilibrium decreases from 6140.550766 to 6086.308139 when premium coefficient is in the interval [0.50, 0.85] and increases from 6086.308139 to 6113.663471 when premium coefficient is in the interval [0.85, 1.50], while OUETR decreases from 24.637857% to 21.566600%. Fig. 2 presents us with a picture of intuition. Based on what we have discussed, we need to address these points:

157

U 6140 6130 6120 6110 6100 6090 6080 6070 6060 6050 6040 6030 6020 6010 6000 5990 O

OUETREquilibrium

BenchmarkEquilibrium

UETR25Equilibrium

0.55 0.65 0.75 0.85 0.95 1.05 1.15 1.25 1.35 1.45

0.5

0.6

0.7

0.8

0.9

1.0

1.1

1.2

1.3

1.4

1.5

t2

Fig. 2. Sensitivity analysis on premium coefficient in benchmark equilibrium, UETR25equilibrium, and OUETR-equilibrium.

(3) No matter how the premium coefficient varies, UETR25 equilibrium is surely inferior to OUETR-equilibrium; (4) No matter how the premium coefficient varies, OUETR equilibrium is surely superior to benchmark equilibrium. 7. Conclusions

(1) The utility function of UETR25 equilibrium is a decreasing curve when premium coefficient grows in the interval of [0.50, 1.50]. And the utility function of OUETR equilibrium is a hyperbola with the lowest point when premium coefficient is 0.85, which means that the curve of UETR25 equilibrium is decreasing when premium coefficient is into the interval [0.50, 0.85], while the curve of UETR25-equilibrium is increasing when premium coefficient is into the interval [0.85, 1.50]; (2) When premium coefficient is into the interval [0.50, 0.65], utility value of benchmark equilibrium is less than that of UETR25-equilibrium, and when premium coefficient is into the interval [0.70, 1.50], utility value of benchmark equilibrium is more than that of UETR25-equilibrium. In most cases benchmark equilibrium is superior to UETR25 equilibrium;

In order to persevere in the policy of reform and opening up and promote the growth of China's economy, China desired financial support from foreign countries. Thus it is natural that FIEs are used to be regarded as the vehicle for bringing foreign investment home. With ample capital and high-quality technology, FIEs did improve China's economic development. In spite of these contributions, conditions have been changed into a different setting. The policy of preferential treatment for FIEs has entangled China with new issues in great need of being settled down. In light of the situations FIEs have caused with the treatment, the central government accomplished the validation of the new Law on Enterprise Tax with the unified tax rate of 25%. Inspired by the consequent impacts of the tax policy reform and motivated to the sparse literature that has deeply explored the optimal

Table 7 Sensitivity analysis on premium coefficient in benchmark equilibrium, UETR25-equilibrium, and OUETR-equilibrium. Premium coefficient

Utility value: benchmark equilibrium

Utility value: UETR25 equilibrium

OUETR equilibrium Utility value

OUETR (%)

0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.35 1.40 1.45 1.50

6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742 6084.5742

6139.7882 6118.9946 6101.6665 6087.0393 6074.5522 6063.7857 6054.4205 6046.2097 6038.9541 6032.5172 6026.7585 6021.5837 6016.9110 6012.6729 6008.8130 6005.2843 6002.0470 5999.0674 5996.3165 5993.7697 5991.4055

6140.5508 6122.3884 6109.3356 6099.9608 6093.5963 6089.5491 6087.2641 6086.3081 6086.3506 6087.1421 6088.4955 6090.2702 6092.3595 6094.6823 6097.1763 6099.7934 6102.4961 6105.2551 6108.0475 6110.8550 6113.6635

24.637857 24.064926 23.335238 22.931068 22.611002 22.366942 22.185205 22.050999 21.951406 21.876409 21.818725 21.773223 21.736336 21.705594 21.679277 21.656182 21.635463 21.616518 21.598917 21.582350 21.566593

158

J. Ji et al. / Economic Modelling 33 (2013) 149–158

unified enterprise tax rate for China, this paper investigates it for China's manufacturing sectors and examines the effect of enterprise tax reform on economy system. We conduct through a general equilibrium model for production economies and the succeeding calibrated parameters based on the data of China in 2007. In the general equilibrium model, to sketch the outlines of supply and demand sides, we set out our production sectors, Agriculture, SOEs, FIEs and OPEs, endowed with Cobb–Douglas production functions, and a representative consumer who is trying to maximize his utility subject to budget constraint. Employing the relevant data of 2007, we harvest necessary parameters in model calibration. When calculating the optimal unified tax rate, we suppose that tax revenue is fixed and let a proportional change exist in net trade. The results show that the optimal unified tax rate is 21.82%. The results above are complemented by completing a sensitivity analysis in which the real enterprise tax rate for FIEs is limited into the interval of [0%, 25%]. Reapplying the algorithms in dealing with the optimal equilibrium, we find that compared with the equilibrium with the unified tax rate of 25%, the representative consumer's utility has improved in the equilibrium with the optimal unified tax rate. The globally optimal enterprise tax rates are studied as well with unrestrainted conditions in the model. The optimization results of globally optimal enterprise tax rates are 33.11%, 18.17% and 18.06% for SOE, FIE and OPE sectors, respectively. The responding utility value has increased beyond that of the benchmark equilibrium. Comparing the benchmark equilibrium, the equilibrium with unified enterprise tax rate of 25% and the equilibrium with globally optimal tax rates, we argue that the optimal unified tax rate is located at around 21.82%, so with tax revenue given the optimal unified tax rate is efficient.

Further to confirm the conclusion, this paper does sensitivity analysis on enterprise tax rate for FIEs and premium coefficient. When the real enterprise tax rate for FIEs increases from 0% to 25%, the optimal unified enterprise tax rate increases from 20.65% to 22.50%; while premium coefficient changes from 0.5 to 1.5, the optimal unified enterprise tax rate varies from 24.64% to 21.57%. The results show that the optimal unified enterprise tax rate is around 21.82%. References Chen, B.L., 2003. Tax evasion in a model of endogenous growth. Review of Economic Dynamics 3, 381–403. Harberger, A.C., 1959. The corporation income tax: an empirical appraisal. Tax Revision Compendium 1 (House Committee on Ways and Means, 86th Congress, First Session), pp. 231–240. Harberger, A.C., 1962. The incidence of the corporation income tax. Journal of Political Economy 70, 215–240. Shoven, J.B., Whalley, J., 1992. Applying General Equilibrium. Cambridge University Press. Shoven, J.B., Whalley, J., 1972. A general equilibrium calculation of the effects of differential taxation of income from capital in the united state. Journal of Public Economics 1, 281–321. van der Hoek, P., Kong, S., Li, Z., 2008. The dual corporate income tax in China: the impact of unification. Public Finance and Management 48, 655–677. Wang, C.Y., 1995. The unification of two income tax policies. Tax Study 35, 7–16. Whalley, J., Wang, L., 2007. The unified enterprise tax and SOEs in China. NBER Working Paper 12899. National Bureau of Economic Research, Inc. Whalley, J., Zhang, S., 2006. State-owned enterprise behaviour responses to trade reforms: some analytics and numerical simulation results using Chinese data. NBER Working Paper 12780. National Bureau of Economic Research, Inc.